• In mixed effect models, you can add a so-called "Random Effect" to your data.
• This categorical covariate represents a combination of all the factors that you cannot measure, but will affect the data largely in the same way
• This is a very useful and important part of any ecological statistical model

Examples

Bees in Week 4 Homework

• For example, in Week 4 Exercise 2, we use the Hive ID as a random effect in our model of the relationship between spore density and infection in honeybee colonies.
  • This is because each hive had multiple observations, but each observation in a hive is more similar to its neighbors than to the other hives
  • So we would expect there to be some correlation

My Ecology group project on white-tailed eagle nests

• For our research proposal project, we planned to use a GLMM to measure changes in white-tailed eagle diet between urban and rural nests. The Nest ID would be used as a random effect, to account for random factors that affect a certain nest differently than its neighbours